Estudia

• CG1 Organize and evaluate the learning and the research activity themselves and develop strategies to improve them.
• CG1- Organize and evaluate the learning and the research activity themselves and develop strategies to improve them
• CB6 Possess and understand the knowledge that provides a basis or opportunity to be original in the development and/or application of ideas, often in a research context.
• CB6- Possess and understand the knowledge that provides a basis or opportunity to be original in the development and/or application of ideas, often in a research context
• CB7 That students know how to apply the knowledge acquired and their ability to solve problems in new or unfamiliar environments within broader contexts related to their area of ??study.
• CB7- That students know how to apply the knowledge acquired and their ability to solve problems in new or unfamiliar environments within broader contexts related to their area of ??study
• CB10 That students have the learning skills to allow them to continue studying in a way that will mostly be self-directed or autonomous.
• CB10- That students have the learning skills to allow them to continue studying in a way that will mostly be self-directed or autonomous
• CE1 Programming, at an advanced level, in the languages and libraries most used in intelligent field robotics.
• CE1- Programming, at an advanced level, in the languages and libraries most used in intelligent field robotics
• CE2 Analyse a problem related to intelligent autonomous systems and identify the appropriate techniques and tools to solve it.
• CE2- Analyse a problem related to intelligent autonomous systems and identify the appropriate techniques and tools to solve it
• CE6 Know and understand when and how to use the main sensors and actuators available for intelligent field robots.
• CE6- Know and understand when and how to use the main sensors and actuators available for intelligent field robots
• CE7 Understand and be able to apply the main computer-based perception techniques.
• CE7- Understand and be able to apply the main computer-based perception techniques
• CE8 Understand the mathematical foundations of intelligent robotic system algorithms.
• CE8- Understand the mathematical foundations of intelligent robotic system algorithms

• Apply techniques of modelling and calibrating computer vision systems.
• Compute 3D information of the real world from 2D image projections
• Apply the principles of triangulation, stereovision, and multicamera geometry
• Understand the limitations of some feature detecting and feature matching algorithms and how to remove false data associations
• Basic working knowledge of Structure-from-Motion and Visual Odometry
• Building creative proposals

1. Introduction

          1.1. Course organization: Objectives, Overview, Contents, Bibliography, Evaluation, Practical Sessions

2. Basic concepts of Projective Geometry in Computer Vision

          2.1. Linear Algebra

          2.2. Points and vectors

          2.3. Translations and Rotations

          2.4. Homogeneous Coordinates

          2.5. Inverses and Transposes

3. Image formation and Camera Modelling

          3.1. Optical Sensors

          3.2. The pinhole model

          3.3. Intrinsic and extrinsic parameters

          3.4. Computing the calibration matrix

          3.5. Effect of camera lenses

4. Image Primitives

          4.1. Interest point detectors

          4.2. Harris and Hessian detectors

          4.3. Similarity measures: SAD, SSD, Correlation

          4.4. Introduction to Scale invariant features

5. Feature detectors and descriptors

          5.1. Feature detectors

          5.2. Invariance

          5.3. Descriptors

          5.4. Review of SIFT

6. Robust Estimation in Computer Vision

          6.1. Probabilistic methods

          6.2. Computing the homography matrix

          6.3. Outlier rejection: Random Sampling Consensus

          6.4. Applications: Planar motion estimation, Mosaicing, etc.

7. Multiple view geometry

          7.1. The principle of Triangulation

          7.2. Stereo vision

          7.3. Epipolar geometry

          7.4. Computing the Fundamental matrix

          7.5. Trinocular constraints and n-camera constraints

8. Structure-from-Motion

          8.1. Review of SfM approaches

          8.2. Main components of 3D model creation pipeline

9. Real-time Computer Vision and Vision applied to Robotic systems

          9.1. Visual odometry

          9.2. Incremental approaches and visual SLAM

          9.3. Review and examples of applied to field Robotics

10. Non-conventional optical imaging systems

          10.1. Omnidirectional vision systems

          10.2. Multispectral and hyperspectral

          10.3. Event-based cameras, range gating and others

• Hartley, Richard (2003). Multiple view geometry in computer vision (2nd ed.). Cambridge [etc.]: Cambridge University Press.
• Ma, Yi (2004). An Invitation to 3-D vision : from images to geometric models. New York [etc.]: Springer, cop..

Assessment activities:

About evaluation:

50% Exam.
50% Practical Sessions (Every exercise Mark >= 3/10)

Specific criteria for the "No show" grade:
Lab assignments are mandatory. Failure to deliver a lab assignment implies that the student will not be evaluated in the module.

Single Assessment:
Exam of theoretical and practical contents of the subject. In order to be able to do this, it will be necessary to first deliver two alternative labs that will be provided to students who opt for the single assessment.

The final grade will be 80% of the exam and 20% of the labs.

If deemed necessary, a meeting will be organized where teachers can ask questions they deem appropriate about the lab reports delivered.

For the students to be elegible for the single assessment, they should apply within the deadlines set and in accordance with the procedures and criteria established by the Governing Board of the center.

Minimum requirements to pass:
To pass the module, the global mark must be >= 5/10

Students can arrange tutorial sessions with the professor by contacting the professor via email. Whenever possible, questions and doubts will be solved via email. Otherwise the tutorial will be conducted using Google Meet or face-to-face.

Presentation of information about the course and course activities will be done through Moodle. Google Meet or Teams will be used for non-contact sessions. All message communication between the professors and the students will be made by internal Moodle messaging system or by email. Students will use Moodle to upload reports.

Students should be familiar with programming in Matlab. Additionally students should have a basic level of programming in C++ and Python.